A Better AI.
- Thread starter Blake
- Start date
DrJambo
Crash-test dummy
It's strange, but I always thought that the AI needed improvements to the military aspect far more than to its ability to speed through the tech tree...
I am a Monarch level player and I usually find I'm behind in tech in the mid game. I start off well and finish well (mainly because I've beaten up on some AI by then).
I've never liked the way the AI will freely tech trade with other AI while at the same time they'll be extremely reluctant to partake in good deals with the human. My hope is that one day the AI will be more adept at warring and be more likely to instigate inter-AI wars, creating more AI-AI vassalizations and thus greater obstacles for the human.
Most games I see now (in 2.08) seem to be AI builder affairs, which are rather boring tech races. If you as the human don't go to war, then you can forget about beating the AI in a builder game at the harder levels. The thing is, the improvement to the AI tech rate is making players drop levels, whereas the military game hasn't really changed at all. In my case I was a Monarch player. With 2.08 it's now extremely hard to keep up with the AI science rate at Monarch, but if I drop to Prince, the military game is too easy....
I am a Monarch level player and I usually find I'm behind in tech in the mid game. I start off well and finish well (mainly because I've beaten up on some AI by then).
I've never liked the way the AI will freely tech trade with other AI while at the same time they'll be extremely reluctant to partake in good deals with the human. My hope is that one day the AI will be more adept at warring and be more likely to instigate inter-AI wars, creating more AI-AI vassalizations and thus greater obstacles for the human.
Most games I see now (in 2.08) seem to be AI builder affairs, which are rather boring tech races. If you as the human don't go to war, then you can forget about beating the AI in a builder game at the harder levels. The thing is, the improvement to the AI tech rate is making players drop levels, whereas the military game hasn't really changed at all. In my case I was a Monarch player. With 2.08 it's now extremely hard to keep up with the AI science rate at Monarch, but if I drop to Prince, the military game is too easy....
Yeah that'd be fine as long as you give full credit (yeah all that usual jazz).TAfirehawk said:
Okay on the topic of AI Cultural Victory, it'd be pretty easy for the human to bust an AI cultural victory since all it takes is capturing and burning one of the lichpin cities and that's the end of the cultural victory. The AI needs to have MUCH better concepts of defense and such before it'll be able to avoid losing a big 3 city to attack. Even isolation isn't a great defense, since it mainly works for the human because the AI is dumb in the ways of astronomy, a human would easily sail over a strike-force and end the victory. A cultural victory is also a game-long plan, something the AI doesn't do at all well.
DrewBledsoe
Veteran QB
Blake said:
I don't think it possible for an AI to ever achieve a cultural victory, for the reasons you mentioned above. As you say, all it takes is for the human to capture / destroy one of the three "culture" cities, and that plan is all over. This will always be possible for a determined enough human player. The only reason at present, that a human player can achieve a cultural victory, is because the AI does nothing to disrupt this.
As the AI becomes better, it too should try to prevent the player from completing a cultural victory (that is if you wish it truly to become "human like" in it's play). Therefore, the logical conclusion is that cultural victory becomes impossible. I've never bothered with cultural victory myself, but to obsolete that option, isn't fair for players who enjoy playing in that style.
Instant_Cereal
Warlord
- Joined
- Nov 27, 2005
- Messages
- 264
=DOCTOR= said:
I'm a monarch player as well and your games seem to mirror mine. I do well at first, then fall behind, then catch up... usually.
But anyway, yes, the single thing that pisses me off more than any other aspect of the AI is the way it trades techs. You're always treated as the redheaded step-child while they all pass each other techs. All it takes is one civ to obtain a tech and e ventually they'll all have via trade. It also sort of ruins the pace of the game. For example, in some of my games the AI has rifling and the scientific method by 1400. I'm not sure if the AI needs improvements in the teching and building department... seems like it's mostly its ability to wage war it needs help with.
TAfirehawk
WoC Team Manager
Instant_Cereal said:
I think many people share your setiments on the AI tech trading "cheat mode".
DrewBledsoe
Veteran QB
Instant_Cereal said:
I agree too, see this thread here on the same subject http://forums.civfanatics.com/showpost.php?p=4723426&postcount=1
player1 fanatic
Fanatic
The thing is, that AI should be good in cultural stuff enough so (if AI is improved), that it could win against player that himself isn't focused on cultural win, or isn't focused on conquering that AI.
Of course, it will never be good enough as human who is focused on his own win, or good enough against human warmonger. But in other instances, I think it could be made good enough to disrupt human player, in similar way as it can do now with Space Race win. That could be one great gameplay improvement.
Of course, it will never be good enough as human who is focused on his own win, or good enough against human warmonger. But in other instances, I think it could be made good enough to disrupt human player, in similar way as it can do now with Space Race win. That could be one great gameplay improvement.
Blake said:
I wonder if the AI would have a better chance at a cultural victory if a more "cumulative" cultural victory condition existed, where the total culture of all the cities triggered the victory. (I'm a little fuzzy on this, but I believe Civ 3 allowed for that kind of cultural victory.) Then it might be more difficult for a human player to prevent a cultural victory. It might also make it more possible for the AI to achieve because the planning necessary for a cultural victory would be simplified. i.e. "Build lots of culture-producing stuff wherever I can." I think a cumulative cultural victory condition would be a nice addition to the game, even if the AI didn't try for it.
By the way, thanks for all the good work you've done on the AI.
TAfirehawk
WoC Team Manager
Pudd'nhead
Chieftain
Blake, I have yet to try your mod (or even the 2.08 patch for that matter, finishing up WOTM2), but you hit the nail right on the head here. For the life of me I can't figure out why anyone would be upset about improving the computer AI, why ego factors such as "not having to drop down a level" trumps the desire for a challenge throught the WHOLE game. I'd rather be a Prince again than sit for hours on end in the tedium of countering the same formulaic half-baked AI response that in the end is little more than a headwind in the face of the human player game's locomotive. In fact these days I hardly play to the very end at all, usually only GOTM that I want to turn in. I usually don't bother because I found after the first few plays that the real challenge is in the early years -- if you get to into the early AD ahead or within striking distance, it seems the rest of the game (the other 80%+ of the playing time!) is just a fait accompli, the last thousand or two years worth is mostly a test of how fast you can get to an already decided conclusion, a slog just to get a final score ranking. Anything that would make the late game has as suspenseful as the early one would be warmly received! Frankly, if that doesn't happen, I'm probably going to lose interest in playing single player civ soon. Though maybe hardcore players who want to max scores, etc, might preferable the predictability & ease of countering the current AI late game, those of us looking for a real GAME, all the way to the END, thirst for the same thing you do.Blake said:
Thanks again, in advance, if these updates are even half of what they seem.
I understand that you contributed to the changes in the 2.08 patch. If you have further improvments, I hope Firaxis will continue to solicit your contributions & will come out with another patch incorporating some of them. It would be very preferable if the changes were part of the standard/vanilla game, rather than a tweak, so that all discussion on the boards, GOTM games etc, would share the same context. Do you know if Firaxis has plans for further Warlords patches?
Thanks again!
drkodos
Emperor
FYI: I am getting blown out of several games mid to late at Emperor level in which I had a lead and thought I had a victory in sight.
It's wonderful.
Several times I have leaders such as Mehmed, or Wang or Churchill be so far behind me at 1000 AD that I thought they were dead in the water only to have them resurrect themselves and eventually give me a beat down.
I just had Tokugawa, of all people, come storming back from the brink of nothingness at 600 AD (last in power out of 10 civs) to be the power & tech (!) leader @ 1700AD in an Emperor level game.
Double
It's wonderful.
Several times I have leaders such as Mehmed, or Wang or Churchill be so far behind me at 1000 AD that I thought they were dead in the water only to have them resurrect themselves and eventually give me a beat down.
I just had Tokugawa, of all people, come storming back from the brink of nothingness at 600 AD (last in power out of 10 civs) to be the power & tech (!) leader @ 1700AD in an Emperor level game.
Double
aelf
Ashen One
Pudd'nhead said:
Because I want to play more balanced game? Because if I want to play an all out aggressive game I'd play some RTS? If people who don't like to have to war for the entire game can only win on Prince, that says a lot about what it takes to master this game.
For me, it's not about not wanting challenge. I still have two very difficult levels to move on to. It's about wanting the potential for a more balanced gameplay to stay.
drkodos
Emperor
Arlborn said:Really? Why? I would lose to see Gandhi going for cultural victory for example. Why cant they make some gp farmer of great artists themselves and some nice production cities for wonders? Is that so hard to progamme the AI to aim more for culture? Are you going to make the AI a warmongger conqueror only cuz its hard to make them go to something else? you gave up of "fair" diplomatic(real diplomatic, not only destrying your enemies before the UN. ould be even nicer if the same AI that is aiming for cultural is also aiming for diplomatic in the same game) and cultural victory for the AI? Well if its true that is a pity then, and Im still looking for some modder to try to improve the AI in that way..Not trying to say anything bad about your actual work, this awesome actually! Just trying to push somebody, ANYBODY, to make the AI goes to any kind of victory, because its really annoying that Firaxis did put so many nice aspects in the game that the AI just dont even try to use. Ah well, still looking
Keep up your work, but please see if there is ANY way to make the AI aims for ALL kind of victories! I hate warmonger games..
I think one reason why the computer would always have a hard time winning culturally is because the human would see it and throw the kitchen sink at them. The only way to do it would be to stack the deck even further with hidden benefits for the game engine.
And, that seems to be the antithesis of the type of "thinking" game I believe is the ultimate goal here.
aelf said:
I agree with this. AI improvements are certainlty good unless they just help the AI civs tech faster. What is the point of that when we have difficulty levels which basically do this already?
It seems the best achieveable result with the work that Blake has done is to improve the AI's performance but then scale back the teching bonuses for the AI, so that it techs at around the same rate.
If you've worked hard to graduate from Noble to Prince it's disheartening to have to go back to Noble, even if Noble is the new Prince.
New version (07), get it on the Source Forge page.
Changes:
This can be summarized as some needed fixes to workers especially for industrial+, some tentative improvements to the tactical side of things and a decent upgrade to the religious aspects. If my changes work as intended you'll also feel a gut-wrenching fear as "Homer" gets popped, however the the target will usually be another AI since AI's tend to found cities more offensively.
The AI should also be more able to hold it's own in border struggles over resources in the early game.
By the way, on the topic of Tech Trading. Try a no tech trading game and you'll see the AI does indeed just tech that fast. "Exclusive" tech trading is not and has never been a crutch the AI uses for faster teching. If you are playing on a "fair" difficulty you should except to be middle of the pack in research much of the time, unless you have a serious edge in research due to terrain and/or wonders.
Changes:
This can be summarized as some needed fixes to workers especially for industrial+, some tentative improvements to the tactical side of things and a decent upgrade to the religious aspects. If my changes work as intended you'll also feel a gut-wrenching fear as "Homer" gets popped, however the the target will usually be another AI since AI's tend to found cities more offensively.
The AI should also be more able to hold it's own in border struggles over resources in the early game.
By the way, on the topic of Tech Trading. Try a no tech trading game and you'll see the AI does indeed just tech that fast. "Exclusive" tech trading is not and has never been a crutch the AI uses for faster teching. If you are playing on a "fair" difficulty you should except to be middle of the pack in research much of the time, unless you have a serious edge in research due to terrain and/or wonders.
There was no balanced gameplay at the higher levels. Warmonger styles became increasingly powerful as you went up the difficulty levels compared to builder styles. As a simple example, consider the traits of the human players leader. While they were all diminished in power, traits such as financial, charismatic and aggresive became stronger relative to other traits while industrious became extremely weak. If the game play at these levels was balanced the relative power of these traits would remain unchanged.
I'm of the opinion that the difficulty levels above noble existed as a means of compensating for the inadequacies of the AI and that they worked by allowing players to exploit these inadequacies. Currently it's the AIs building which has been improved so that it remains inadequate at warfare. As the AI becomes better and better, these levels become more and more relic-like. Had an AI that is, to us, improved been available from the release of Vanilla I question whether the higher difficulties would be present in the game at all. They certainly wouldn't be present in their current form.
Finally, the percieved increased emphasis on warmonger styles at higher difficulty levels would be nothing compared to its increased emphasis had the majority of AI improvements been applied to warfare tactics. While I don't think it's your intention, I suspect that your arguement effectively comes down to a complaint that the AI was improved at all.
There is a strong case to be made for revised difficulty levels but it may be pointless for an official patch to include these until the rate at which AI improvements are added has effectively stopped. Even then, it may take a lot of testing to develop them. Or not. A new set of difficulty levels could simply be based on increased health, happiness and free military upkeep.
